Simple theory predicts that the imaging time or injected dose in a PET study can be reduced by roughly an order of magnitude by using time-of-flight (TOP) information. This reduction can be obtained by improving the coincidence timing resolution, and so would be achievable in clinical, whole body studies using with PET systems that differ little from existing cameras. The purpose of this proposal is to realize the best possible timing resolution in a prototype PET camera and verify the performance improvement experimentally, measuring how the improvement depends on the timing resolution. The potential impact of this development is large, especially for oncology studies in large patients, where it is sorely needed. Time-of-flight PET was extensively studied in the 1980's but died in the 1990's, as it was impossible to reliably achieve sufficiently good timing resolution without sacrificing other PET performance aspects. Recent improvements in technology have renewed interest in TOP PET, but changes in the last decade in both technology and the medical uses for PET mean that the true benefits of TOP in modern clinical PET are not well understood. By quantifying these benefits as a function of timing resolution, we will guide future commercial and academic TOP PET camera designers by allowing them to make informed choices. To realize this goal, we will construct a single ring "demonstration" PET camera with LSO scintillator crystals that achieves 300 ps coincidence timing resolution. This is a factor of two better timing resolution than any LSO-based PET camera ever constructed or currently proposed, and the simple estimates predict that it would reduce the noise variance by a factor of 8 in whole body imaging. Clinically, this variance reduction could reduce image noise, decrease the injected dose, shorten the imaging time, or provide a combination of these benefits. We will quantify the improvement in noise as a function of timing resolution using a combination of simulation, phantom, and human subject studies, including the clinically relevant tasks of tumor detection and staging. We will then rebuild the camera with a newly developed scintillator (Lul3:Ce) that has exceptional PET performance (high sensitivity, 3% energy resolution, and 125 ps timing resolution) and repeat these measurements. This timing resolution is over three times better than any TOP camera ever built or proposed, and is predicted to give an enormous (18-fold) noise variance reduction.